Intraventricular hemorrhage (IVH) or hemorrhage into the germinal matrix tissues of the developing brain, remains a major problem of preterm neonates. We enrolled 505 neonates of 600 - 1250 g birth weight in a prospective, randomized, placebo-controlled multicenter trial to test the hypothesis that low dose indomethacin would lower the incidence and decrease the severity of IVH. The 431 infants with no evidence for IVH at 6 - 11 hours of age were admitted to the formal trial. There were no differences in the birth weight, gestational age, sex, Apgar scores and percent of infants treated with surfactant between the indomethacin and placebo groups. Twenty five (12%) indomethacin treated and 40 (18%) placebo treated infants developed IVH (p=0.03, trend test). One indomethacin patient experienced parenchymal involvement of hemorrhage compared to 10 placebo infants (p-0.01). Sixteen indomethacin infants and 29 control infants died (p=0.08), and there was a difference favoring indomethacin with respect to survival time (p-0.06). There were no differences in adverse events. Neurodevelopmental follow-up assessments performed at 7, 18 and 36 months' corrected age (C.A.) on all patients document 88% follow-up at all age intervals and demonstrate significantly higher Bayley mental and motor scores for those infants with no IVH compared to those with hemorrhage. In addition, those infants who had received indomethacin tended to perform better on the Stanford Binet at 36 mo. C.A. (p=0.11). We hypothesize that those children randomized to indomethacin will perform significantly better on measures of cognitive function, motor abilities, and attention that those infants randomized to saline placebo. This outcome reflects both our documented evidence indicating that indomethacin lowers the incidence and severity of IVH in very low birth weight infants and our preliminary data which demonstrate that indomethacin has no adverse effect on cognitive outcome at 7, 18, and 36 mo. C. A. We propose to serially study measures of IQ, attention, executive function, motor skills, behavior, and school readiness in our study population at both 54 and 72 mo. C. A. to assess the influence of indomethacin on these domains. Our secondary specific aims will study the influence of IVH on developmental outcome at 54 and 72 mo. C. A. and mMultivariate analyses will be performed to determine independent and important predictors of outcome at these two ages.